Volume 22, Issue 1, Pages 111-119.e4 (July 2017) HIV-1-Mediated Downmodulation of HLA-C Impacts Target Cell Recognition and Antiviral Activity of NK Cells  Christian Körner, Camille R. Simoneau, Philipp Schommers, Mitchell Granoff, Maja Ziegler, Angelique Hölzemer, Sebastian Lunemann, Janet Chukwukelu, Björn Corleis, Vivek Naranbhai, Douglas S. Kwon, Eileen P. Scully, Stephanie Jost, Frank Kirchhoff, Mary Carrington, Marcus Altfeld  Cell Host & Microbe  Volume 22, Issue 1, Pages 111-119.e4 (July 2017) DOI: 10.1016/j.chom.2017.06.008 Copyright © 2017 Elsevier Inc. Terms and Conditions

Cell Host & Microbe 2017 22, 111-119. e4DOI: (10. 1016/j. chom. 2017 Copyright © 2017 Elsevier Inc. Terms and Conditions

Figure 1 HIV-1 Clones Differentially Modulate HLA-C Expression after In Vitro Infection Effects of HIV-1 infection on HLA-C surface expression on primary CD4+ T cells as determined by flow cytometry. (A) HLA-C levels of infected and uninfected cells after infection with NL4-3 and JR-CSF. (B) Extent of HIV-1-mediated HLA-C downmodulation across various HIV-1 clones as determined by the ratio between HLA-C expression (median fluorescence intensity, MFI) of uninfected and infected cells (n = 6). Boxplots define median and IQR. (C) Baseline HLA-C expression of uninfected CD4+ T cells (clear) and after infection with JR-CSF (red) (n = 24). Black bars represent the median. (D) Correlation analysis of HLA-C levels (MFI) between uninfected and JR-CSF-infected CD4+ T cells. (E) Correlation analysis between baseline HLA-C expression levels of uninfected CD4+ T cells and the extent of HIV-1-mediated HLA-C downmodulation. Statistical analyses were as follows: Wilcoxon matched-pairs signed-rank test and Spearman rank correlation. See also Figure S3. Cell Host & Microbe 2017 22, 111-119.e4DOI: (10.1016/j.chom.2017.06.008) Copyright © 2017 Elsevier Inc. Terms and Conditions

Figure 2 HIV-1 Infection Alters Binding of KIR2DL1 and KIR2DL3 to HLA-C KIR binding to HIV-1-infected and uninfected CD4+ T cells measured by flow cytometry and stratified by donor HLA-C haplotypes (C2Hom, n = 9; C1/2, n = 11; C1Hom, n = 12). (A and B) KIR binding to infected cells after infection with JR-CSF (A, KIR2DL1, blue; B, KIR2DL3, yellow) as compared to uninfected cells (clear) displayed as relative fluorescence intensity. Black bars represent the median. (C) Correlation analysis between HIV-1-associated reduction of HLA-C expression and KIR binding on infected cells given as fold change reduction compared to uninfected cells. KIR2DL1 binding to cells from C2Hom donors (n = 9) is displayed in blue; KIR2DL3 binding to cells from C1Hom donors (n = 12) in yellow. Statistical analyses were as follows: Wilcoxon matched-pairs signed-rank test and Spearman rank correlation. See also Figures S1 and S3. Cell Host & Microbe 2017 22, 111-119.e4DOI: (10.1016/j.chom.2017.06.008) Copyright © 2017 Elsevier Inc. Terms and Conditions

Figure 3 HLA-C-Licensed KIR2DL+ NK Cells Display Impaired Antiviral Activity (A) Assessment of HLA-C-mediated licensing of NK cells determined by frequency of CD107a expression and TNF-α production (left panel). Frequency of CD107a+ (middle panel) and TNF-α-producing cells (right panel) among different NK cell subsets (n = 12; C2Hom, n = 4; C1Hom, n = 8). (B) Antiviral capacity of NK cell subsets. KIR2DL1+ and KIR2DL3+ NK cells were sorted by fluorescence-activated cell sorting (FACS) (left panel). Viral loads in the presence or absence of bulk NK cells (middle panel). Level of NK cell-mediated inhibition of HIV-1 replication exhibited by NK cell subsets as compared to CD4+ T cells alone (right panel). All donors were KIR haplotype A to avoid confounding effects of activating KIR and KIR2DL2 and homozygous for either HLA-C group (C1Hom, n = 6, C2Hom, n = 2). (A and B) Licensed (2DL+ Self) and unlicensed KIR2DL+ NK cells (2DL+ Nonself) are defined by the expression of KIR2DL1 and KIR2DL3 and the donor HLA-C haplotype (2DL+ Self: C2Hom, KIR2DL1+/L3(−); C1Hom, KIR2DL3+/L1(−); 2DL+ Nonself: C2Hom, KIR2DL1(−)/L3(+); C1Hom, KIR2DL3(−)/L1(+)). Bulk and KIR2DL1(−)/L3(−) NK cells (2DL(−)) served as internal controls. (C) Antiviral capacity of bulk NK cells in the presence or absence of KIR2DL antibodies (left panel; n = 17; C1Hom, n = 5; C1/C2, n = 9; C2Hom, n = 3). Viral loads in the presence or absence of NK cells (middle panel) or in the presence of KIR antibodies that do (α2DL Self) or do not (α2DL Nonself) bind to donor HLA-C ligands (right panel). α2DL Self indicates addition of α2DL1 to cells from HLA-C2+ donors or α2DL3 to cells from HLA-C1+ donors; α2DL Nonself indicates addition of α2DL1 to cells from HLA-C1Hom donors or α2DL3 to cells from HLA-C2Hom donors. Black bars represent the median. Statistical analyses were as follows: Wilcoxon matched-pairs signed-rank test. See also Figures S2 and S3. Cell Host & Microbe 2017 22, 111-119.e4DOI: (10.1016/j.chom.2017.06.008) Copyright © 2017 Elsevier Inc. Terms and Conditions

Figure 4 NK Cells Are Able to Sense HIV-Mediated Changes in HLA-C Expression (A) Comparison between JR-CSF WT and the Vpu mutant strain (Vpu mut) in terms of their ability to downmodulate HLA-C. Vpu mut contained the Vpu protein sequence of NL4-3 in positions 4 and 5 (L4I+Q5del). Relative HLA-C expression of CD4+ T cells infected with either WT (clear) or Vpu mut (red) as compared to uninfected cells, given as fold change (n = 8). HLA-C expression was quantified as median fluorescence intensity. (B) KIR2DL binding (KIR2DL1 and KIR2DL3) to CD4+ T cells infected with either WT (clear) or Vpu mut (red) as compared to uninfected cells, given as fold change (n = 11; KIR2DL1, n = 6; KIR2DL3, n = 5). KIR binding was quantified as relative fluorescence intensity. (C) Levels of NK cell-mediated inhibition of HIV-1 replication after 7 day co-culture with CD4+ T cells infected with either WT or Vpu mut (n = 8). Inhibition of HIV-1 replication was calculated from viral loads (copies/mL) in the presence or absence of NK cells. (D) Frequency of CD107a+ NK cells after 6 hr co-incubation with various target cells. NK cells single positive (sp) for KIR2DL1, KIR2DL3, KIR3DL1, and NKG2A were identified by Boolean gating. Licensed KIR2DL+ NK cells were defined as KIR2DL1sp NK cells from C2+ donors (n = 3) or KIR2DL3sp NK cells from C1+ donors (n = 8) (gray boxplots). Unlicensed NK cells lacked all mentioned inhibitory receptors (white boxplots). (E) HIV-specific responses of licensed KIR2DL+ NK cells (n = 11) after exposure to CD4+ T cells infected with either WT (clear) or Vpu mut (red). Absolute HIV-specific responses are given as percent CD107a+ NK cells after adjustment for non-HIV-1-specific responses (NK cells alone and mock-infected CD4+ T cells). (F) Comparison of the relative HIV-specific responses of licensed KIR2DL+ NK cells after exposure to CD4+ T cells infected with either WT (clear) or Vpu mut (red). Relative HIV-1-specific responses are given as fold change compared to HIV-specific responses of unlicensed NK cells. Black bars represent the median. Boxplots represent median and IQR. Statistical analyses were as follows: Wilcoxon matched-pairs signed-rank test. See also Figure S3. Cell Host & Microbe 2017 22, 111-119.e4DOI: (10.1016/j.chom.2017.06.008) Copyright © 2017 Elsevier Inc. Terms and Conditions